spark-optimization
Pass
Audited by Gen Agent Trust Hub on Sep 30, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill provides templates for interacting with external data sources (e.g., S3 buckets, Spark tables) and processing user-supplied column names and schemas. This creates a surface where maliciously crafted data or metadata could influence downstream processing.
- Ingestion points: S3 data paths, table names, and column identifiers used in Spark transformations in SKILL.md.
- Boundary markers: The instructions do not specify delimiting markers for user-supplied strings or data content.
- Capability inventory: The skill utilizes capabilities for reading from and writing to S3 buckets and performing network-intensive shuffle operations as seen in Pattern 1 and Pattern 6.
- Sanitization: No explicit sanitization or validation of input data paths or column names is implemented in the provided code patterns.
Audit Metadata